Article Text

Download PDFPDF

Disease progression of non-alcoholic fatty liver disease: a prospective study with paired liver biopsies at 3 years
  1. Vincent Wai-Sun Wong1,2,
  2. Grace Lai-Hung Wong1,2,
  3. Paul Cheung-Lung Choi3,
  4. Anthony Wing-Hung Chan3,
  5. Mia Ka-Po Li1,2,
  6. Hoi-Yun Chan1,2,
  7. Angel Mei-Ling Chim1,2,
  8. Jun Yu1,2,
  9. Joseph Jao-Yiu Sung1,2,
  10. Henry Lik-Yuen Chan1,2
  1. 1Department of Medicine and Therapeutics, The Chinese University of Hong Kong
  2. 2Institute of Digestive Disease, The Chinese University of Hong Kong
  3. 3Department of Anatomical and Cellular Pathology, The Chinese University of Hong Kong
  1. Correspondence to Professor Henry L-Y Chan, Department of Medicine and Therapeutics, 9/F, Prince of Wales Hospital, 30-32 Ngan Shing Street, Shatin, Hong Kong; hlychan{at}cuhk.edu.hk

Abstract

Background Patients with non-alcoholic steatohepatitis (NASH) have increased mortality and liver-related complications. In contrast, simple steatosis is considered benign and non-progressive.

Objective To investigate disease progression in patients with different degrees of non-alcoholic fatty liver disease (NAFLD) activity.

Design Prospective longitudinal hospital-based cohort study.

Patients Fifty-two patients (age 44±9 years) with biopsy-proven NAFLD had liver biopsies repeated at month 36.

Results Among 13 patients with simple steatosis at baseline, 2 (15%) had a normal liver at month 36, 3 (23%) continued to have simple steatosis, 5 (39%) developed borderline NASH and 3 (23%) developed NASH. Among 22 patients with borderline NASH at baseline, 4 (18%) had simple steatosis and 13 (59%) had borderline NASH at month 36, while 5 (23%) developed NASH. Among 17 patients with NASH at baseline, 10 (59%) continued to have NASH and 6 (35%) had borderline NASH at month 36. Only 1 (6%) patient regressed to simple steatosis. Overall, 14 (27%) patients had fibrosis progression, 25 (48%) had static disease, and 13 (25%) had fibrosis regression. Reduction in body mass index and waist circumference was independently associated with non-progressive disease activity and fibrosis. The baseline serum levels and month 36 changes in adiponectin, tumour necrosis factor α, interleukin 6 and leptin were not associated with disease progression. Serum cytokeratin-18 fragment level reflected disease activity and its change correlated with the change in NAFLD activity score (R=0.51, p<0.001).

Conclusions Patients with simple steatosis may still develop NASH and fibrosis progression. Weight reduction is associated with non-progressive disease. All patients with NAFLD should undergo periodic assessment and lifestyle modification.

View Full Text

Statistics from Altmetric.com

Significance of this study

What is already known about this subject?

  • Non-alcoholic steatohepatitis is a progressive disease that may result in cirrhosis and liver-related complications.

  • Simple steatosis is considered benign and non-progressive, but longitudinal data are largely lacking.

  • Dysregulated adipokines are associated with insulin resistance and inflammation. It is unclear if adipokines affect progression of non-alcoholic fatty liver disease (NAFLD).

What are the new findings?

  • Fifty-eight per cent of patients with histological NAFLD activity score <3 had increased activity score at 3 years, and 28% had fibrosis progression.

  • Reduction in body mass index and waist circumference were independent factors associated with non-progressive disease.

  • Serum cytokeratin-18 fragment level reflected disease activity. Its change had moderate correlation with the change in NAFLD activity score.

How might they impact on clinical practice in the foreseeable future?

  • Since patients with NAFLD with different disease activity may still have disease progression, periodic assessment is necessary in all patients with NAFLD.

  • Our study also highlights that weight reduction is central in the prevention of disease progression.

Non-alcoholic fatty liver disease (NAFLD) is one of the most common liver diseases world wide.1 It may progress to cirrhosis, liver failure and hepatocellular carcinoma. It is also strongly associated with metabolic syndrome and cardiovascular risk factors.2 3 As a result, patients with NAFLD have increased mortality and morbidity compared with the general population.4–6 Malignancy, ischaemic heart disease and liver-related complications are the leading causes of death in patients with NAFLD.4

Based on histological disease activity, NAFLD can be divided into simple steatosis and non-alcoholic steatohepatitis (NASH). Increased mortality and liver-related complications are mainly seen in patients with NASH.7 On the other hand, simple steatosis is believed to be benign and non-progressive. However, this claim is limited by the paucity of data in this area. In a recent systematic review, only 10 studies comprising 221 patients with paired liver biopsies were identified from 1989 to 2006.8 Additionally, the intervals between the biopsies were variable, making interpretation of disease progression difficult.7 9–17 Understandably, since the follow-up biopsies were not planned a priori, there was selection bias with over-representation of patients who had more advanced disease. A true prospective study with scheduled paired biopsies and low dropout rate is required.

Another major finding of the systematic review by Argo and colleagues was that factors associated with severe NAFLD in cross-sectional studies might not be associated with disease progression in studies with paired liver biopsies.8 In cross-sectional studies, metabolic factors associated with advanced disease might be the result rather than the promoting factors of disease progression.18

Adipokines are bioactive proteins secreted by adipose tissue. Commonly studied adipokines include leptin, adiponectin, tumour necrosis factor α, interleukin 6 and resistin. The adipokines have important effects on insulin resistance and inflammation. Dysregulated production of adipokines is commonly seen in NAFLD and is associated with disease severity in cross-sectional studies.3 19 Adipokines also affect lipid metabolism and insulin resistance in patients with viral hepatitis.20 However, the effects of adipokines on NAFLD progression have not been studied in longitudinal studies.

In this prospective study with paired liver biopsies 3 years apart, we studied the disease progression in patients with NAFLD with different baseline disease activities. Factors associated with disease progression and the role of adipokines were also studied. In addition, the possibility of predicting disease progression with non-invasive tests of hepatic necroinflammation and fibrosis was explored.

Patients and methods

In 2006, we reported a multicentre cross-sectional study of 80 patients with NAFLD with liver biopsies and adipokine measurements.3 Subjects in one of the participating centres (Prince of Wales Hospital, Hong Kong) were invited to participate in a prospective study. We included patients aged ≥18 years with biopsy-proven NAFLD. We excluded men who consumed more than 20 g of alcohol per day and women who consumed more than 10 g/day using standard questionnaires. Patients with positive hepatitis B surface antigen or anti-hepatitis C virus antibody, antinuclear antibody titre >1/160 or histological features of other liver diseases were excluded. Patients with secondary causes of liver steatosis (eg, use of systemic corticosteroids) were also excluded. The study protocol was approved by the local ethics committee. All patients provided informed written consent.

Follow-up assessment

Patients in the prospective study were followed up every 6 months. At baseline, the patients received one session of dietary counselling from dieticians. They were also encouraged to increase physical activity to at least three times a week, 30 min per session. During each visit, anthropometric measurements including body weight, body height and waist circumference were made. Body mass index (BMI) was calculated as weight (kg) divided by height (m) squared. Waist circumference was measured at a level midway between the lower rib margin and iliac crest with the tape all around the body in the horizontal position. The patients were instructed to fast for 8 h for blood tests including liver biochemistry, glucose, lipids and insulin. Adipokines and cytokeratin-18 fragment were measured at baseline and month 36 using serum stored at −80°C in a refrigerator, which was collected from all patients. Insulin resistance was calculated by the homoeostasis model assessment (HOMA-IR), which was equal to fasting insulin (mU/l)×fasting glucose (mmol/l)/22.5. Adiponectin, tumour necrosis factor α and interleukin 6 were measured by the Quantikine immunoassay (R&D Systems, Minneapolis, Minnesota, USA). Insulin level (Dako, Glostrup, Denmark) and leptin (Diagnostic Systems Laboratory, Webster, Texas, USA) were measured by ELISA. Cytokeratin-18 fragment was measured with the M30-Apoptosense ELISA Kit (Alexis Biochemicals, Lausen, Switzerland).

Metabolic syndrome was defined according to the ethnic-specific criteria by the International Diabetes Federation, which was modified from the National Cholesterol Education Program, Adult Treatment Panel III Guidelines: any three of the following: (1) central obesity (waist circumference ≥90 cm in men and ≥80 cm in women); (2) triglycerides >1.7 mmol/l; (3) reduced high-density lipoprotein-cholesterol (<1.03 mmol/l in men and <1.29 mmol/l in women); (4) blood pressure ≥130/85 mm Hg and (5) fasting plasma glucose ≥5.6 mmol/l; or receiving treatment for the above metabolic abnormalities.21 Diabetes was diagnosed as fasting plasma glucose ≥7.0 mmol/l or 2 h post-challenge glucose ≥11.1 mmol/l after a 75 g oral glucose tolerance test.

The aspartate aminotransferase (AST)-to-platelet ratio index was calculated as AST (/upper limit of normal)/platelet count (×109/l)×100.22 FIB-4 was calculated as age×AST (U/l)/platelet count (×109/l)×√ALT (U/l), where ALT is alanine aminotransferase.23 The NAFLD fibrosis score was calculated according to the following formula: −1.675+0.037×age (years)+0.094×BMI (kg/m2)+1.13×IFG/diabetes (yes=1, no=0)+0.99×AST/ALT ratio–0.013×platelet (×109/l)–0.66×albumin (g/dl), where IFG = impaired fasting glucose.24 The BARD score was the weighted sum of three variables (BMI ≥28=1 point, AST/ALT ratio ≥0.8=2 points, diabetes=1 point).25

Histological assessment

Percutaneous liver biopsy was performed at baseline and month 36 using a 16-gauge Temno needle (Cardinal Health, McGaw Park, Illinois, USA). At study completion, all histological slides were read by two pathologists (PCLC, AWHC) who were unaware of the clinical data and the sequence of the specimens. For specimens with discrepancies in histological scoring, a consensus was reached after discussion between the two pathologists. Liver histology was reported by the pathologists′ global assessment and a semiquantitative scoring. The global assessment was modified from the original description by Matteoni and colleagues.26 Simple steatosis was diagnosed for specimens with fatty liver alone without necroinflammation. Borderline NASH was diagnosed for specimens with lobular inflammation but not to the degree of NASH. NASH was diagnosed for specimens with hepatocytes ballooning or intralobular hepatocyte necrosis. Histological grading and staging of NAFLD were scored according to the system reported by Kleiner and colleagues.27 NAFLD activity score (NAS) was the sum of steatosis, lobular inflammation and hepatocellular ballooning scores. NAS ranged from 0 to 8. For analysis, patients were grouped according to disease activity into NAS <3, 3–4 and ≥5. Fibrosis was staged from 0 to 4, with stage 0=no fibrosis, 1=perisinusoidal or periportal fibrosis, 2=perisinusoidal and portal/periportal fibrosis, 3=bridging fibrosis and 4=cirrhosis.

Statistical analysis

Statistical tests were performed using the Statistical Package for Social Sciences version 16.0. Continuous variables were expressed as mean±SD or median (IQR), as appropriate. Changes in continuous variables were assessed by paired t test or Wilcoxon rank-sum test. Variables between patients with and without increased NAS and with and without fibrosis progression were compared using unpaired t test, Mann–Whitney U test, χ2 test or Fisher exact test as appropriate. Binary logistic regression analysis was performed to identify factors associated with increased NAS and fibrosis progression. The multivariate analysis included significant factors in the univariate analysis. Correlation between changes in histological parameters and changes in non-invasive models was tested using the Spearman test. A two-sided p value of <0.05 was taken as statistically significant.

Results

Among 80 patients in the initial study, 54 patients were from the Prince of Wales Hospital and were eligible for the prospective study. The patient characteristics of the two participating centres were similar. One patient refused to repeat liver biopsy. Follow-up liver biopsy failed in another patient, and he also refused to repeat the procedure. Fifty-two patients with paired liver biopsies were included in the final analysis.

The clinical characteristics at baseline and month 36 are shown in table 1. During 3 years of follow-up, there was slight reduction in serum ALT and insulin level. On the other hand, there was no significant change in anthropometric parameters and lipid profile. Six (12%) and eight (15%) patients developed incident diabetes mellitus and hypertension, respectively. The number of patients with metabolic syndrome was 35 (67%) at baseline and 36 (69%) at month 36. There was no significant difference in the use of oral hypoglycaemic agents and insulin at the two time points.

Table 1

Clinical characteristics of 52 patients with non-alcoholic fatty liver disease (NAFLD) at baseline and month 36

Histological progression

The length of liver biopsy samples was 18±4 mm. In the overall population, there was mild increase in steatosis and ballooning, but reduction in lobular inflammation (table 1). At baseline, 13 (25%), 22 (42%) and 17 (33%) patients had simple steatosis, borderline NASH and NASH, respectively. At month 36, two (4%), eight (15%), 24 (46%) and 18 (35%) patients had normal liver, simple steatosis, borderline NASH and NASH, respectively. Among 13 patients with simple steatosis at baseline, three (23%) continued to have simple steatosis at month 36, five (39%) developed borderline NASH and three (23%) developed NASH. The remaining two patients (male; aged 39 and 48 at baseline) had normal liver histology at month 36, with steatosis found in <5% of hepatocytes, no necroinflammation and fibrosis. Both patients had normal fasting glucose and NAS of 1 point at baseline. At month 36, their waist circumference had reduced by 20 cm and 6 cm, and BMI had reduced by 0.6 kg/m2 and 0.8 kg/m2, respectively. Among 22 patients with borderline NASH at baseline, four (18%) had simple steatosis and 13 (59%) had borderline NASH at month 36, while five (23%) developed NASH. Among 17 patients with NASH at baseline, 10 (59%) continued to have NASH and six (35%) had borderline NASH at month 36. Only one (6%) patient regressed to simple steatosis.

At baseline, 29 (56%), 18 (35%) and five (10%) patients had NAS <3, 3–4 and ≥5, respectively (table 2). At month 36, 17 (33%), 31 (60%) and four (8%) patients had NAS <3, 3–4 and ≥5, respectively. Twenty-six (50%) patients had increased NAS, 11 (21%) had static disease activity and 15 (29%) had reduced NAS from baseline to month 36.

Table 2

Distribution of disease activity at baseline and month 36

The majority of patients with low disease activity had disease progression. Among 29 patients with NAS <3 at baseline, 16 (55%) had score of3–4 and one (3%) had score of 5 points at month 36. The latter patient was a 38 year old woman (NAS increased from 1 to 5). She was receiving dietary treatment for type 2 diabetes. Over 3 years, her BMI rose from 29.2 kg/m2 to 31.6 kg/m2, and waist circumference rose from 96 cm to 99 cm. In contrast, all five patients with NAS ≥5 at baseline had some improvement in NAS at month 36 (table 2).

At baseline, 26 (50%) patients had liver fibrosis, including two (4%) with advanced (F3–4) fibrosis (table 3). At month 36, 24 (46%) patients had liver fibrosis, but seven (14%) patients developed advanced fibrosis (p=0.16 comparing the proportion of patients with advanced fibrosis). Overall, 14 (27%) patients had fibrosis progression by one stage or more, 25 (48%) patients had static fibrosis stage, while 13 (25%) had regression of fibrosis by one stage or more. Five (10%) patients had fibrosis progression by two to four stages.

Table 3

Distribution of fibrosis stage at baseline and month 36

NAS at baseline (R=−0.12, p=0.38) and change in NAS (R=0.23, p=0.098) had poor correlation with the change in fibrosis stage. Progression in fibrosis by one stage or more was seen in 8/29 (28%) patients with NAS <3 at baseline, 5/18 (28%) patients with NAS 3–4 and 1/5 (20%) patients with NAS ≥5 (p=0.94).

Since the length of liver specimen might affect histological interpretation, a sensitivity analysis was performed. Increased NAS was found in 8/14 (57%) patients with biopsy sample length >20 mm and 18/38 (47%) patients with biopsy length <20 mm (p=0.53). Fibrosis progression was found in 4/14 (29%) patients with biopsy length >20 mm and 10/38 (26%) patients with biopsy length <20 mm (p=1.0).

Factors associated with change in NAFLD activity score

Both changes in BMI (R=0.53, p<0.001) and waist circumference (R=0.38, p=0.005) correlated with change in NAS. Patients with increased NAS at month 36 had less reduction in BMI and waist circumference (table 4). They also had less reduction in serum ALT and leptin levels. By multivariate analysis, only the increase in BMI remained as an independent factor associated with worsened NAS (OR for each 1 kg/m2 increase, 1.9; 95% CI 1.2 to 3.1; p=0.012).

Table 4

Factors associated with increased non-alcoholic fatty liver disease (NAFLD) activity score from baseline to month 36

Factors associated with change in fibrosis stage

Both changes in BMI (R=0.28, p=0.049) and waist circumference (R=0.34, p=0.015) correlated with change in fibrosis stage. Patients with fibrosis progression had less reduction in BMI and waist circumference (table 5). They also had higher low-density lipoprotein-cholesterol level and tended to have less reduction in ALT level. By multivariate analysis, increase in waist circumference (adjusted OR for each 1 cm increase, 1.3; 95% CI 1.1 to 1.5; p=0.002) and high baseline low-density lipoprotein-cholesterol (adjusted OR for each 1 mmol/l increase, 2.7; 95% CI 1.2 to 6.1; p=0.019) were independent factors associated with a worsened fibrosis stage.

Table 5

Factors associated with fibrosis progression

Non-invasive assessment of liver injury

Serum cytokeratin-18 fragment level correlated well with NAS both at baseline and month 36. At baseline, serum cytokeratin-18 fragment level of patients with NAS <3, 3–4 and ≥5 was 319 (229–443), 482 (236–1044) and 3045 (310–3924) U/l, respectively (p=0.046 among the three groups). At month 36, serum cytokeratin-18 fragment level of patients with NAS <3, 3–4 and ≥5 was 212 (162–314), 451 (272–645) and 750 (396–2488) U/l, respectively (p<0.001 among the three groups). The change in NAS had moderate correlation with change in serum cytokeratin-18 fragment levels (R=0.51, p<0.001). Patients with increased NAS at month 36 had greater increase in serum cytokeratin-18 fragment levels (table 4).

The change in fibrosis stage only had borderline correlation with change in the AST-to-platelet ratio index (R=0.26, p=0.059), but not with changes in AST/ALT ratio (R=−0.052, p=0.71), FIB4 index (R=0.24, p=0.086), NAFLD fibrosis score (R=0.058, p=0.69) and BARD score (R=−0.071, p=0.63). The change in these non-invasive scores was not significantly different between patients with and without fibrosis progression (table 6).

Table 6

Changes in non-invasive prediction scores of fibrosis in patients with non-alcoholic fatty liver disease (NAFLD) with and without fibrosis progression

Discussion

In this prospective study with paired liver biopsies, patients with simple steatosis commonly had progression in disease activity. Fibrosis progression was seen in 20–30% of patients with both low and high NAS. Weight reduction was the single most important factor associated with static or improved disease activity and fibrosis stage. Non-invasive assessment of fibrosis by biochemical tests was not sufficiently sensitive to predict fibrosis progression.

Compared with previous reports, our study included a relatively large number of patients, was prospectively planned with very low dropout rate and provided detailed metabolic and adipokine profiles. Moreover, patients with simple steatosis were under-represented in previous studies but were also included in this prospective work.7 9–17 Compared with other longitudinal series, the average disease activity of our cohort was milder. While the difference might be partly explained by ethnic and environmental factors, the observation may also be explained by the younger age of our cohort. At baseline, the mean age was 44 years. In comparison, the mean age of other reported longitudinal cohorts ranged from 45 to 58 years.8 It is possible that the disease spectrum would become more severe as the patients become older.

Our study confirmed that simple steatosis was not always quiescent. Twenty-three per cent of patients with simple steatosis developed NASH in 3 years. Additionally, 58% of patients with NAS <3 had increased disease activity after 3 years, and 28% had fibrosis progression. The finding has major clinical implications. Since increasing disease activity may also occur in patients with simple steatosis, regular assessment of disease progression and liver injury is necessary. Owing to its invasive nature, repeated liver biopsies are impractical in routine clinical practice. Further development and refinement of non-invasive tests for NASH and liver fibrosis is required.28 29

Weight reduction was strongly associated with histological improvement or static disease. Even modest weight reduction was beneficial. In patients with static or improved NAS, the mean reduction in BMI and waist circumference was only 0.8 kg/m2 and 3.9 cm, respectively. Similarly, the mean reduction in BMI and waist circumference among patients with static or improved fibrosis stage was 0.5 kg/m2 and 3.3 cm, respectively. Our results concurred with a smaller study of 22 patients with NAFLD, which showed that obesity was the only factor associated with fibrosis progression.15 In lean Korean patients, mild weight gain was associated with incident NAFLD.30 Waist circumference was also associated with liver fibrosis in children and adolescents with NAFLD.31 Although physical activity was not systematically recorded in this study, our results support the notion that weight reduction is beneficial in patients with NAFLD. In an appropriate setting, patients with NAFLD can be motivated to exercise regularly and modify their lifestyle.32 33 In short-term studies, exercise intervention decreased both hepatic and visceral fat as measured by magnetic resonance spectroscopy.34 The future challenge is how to maintain healthy lifestyle and its beneficial effects.35

Recently, a number of clinical models have been developed to predict fibrosis in NAFLD. In general, they have moderate accuracy in diagnosing advanced fibrosis and cirrhosis, with area under the receiver operating characteristics curve typically in the range of 0.6 to 0.9.23 29 However, there is substantial overlap in the scores of patients with and without advanced disease.28 Additionally, a significant proportion of patients are in the gray zone, and their fibrosis stage is undetermined. As such, this study showed that the commonly used scores had limited sensitivity in detecting changes in fibrosis stage with time. On the other hand, direct biomarkers of fibrosis and fibrogenesis, such as hyaluronic acid and the European Liver Fibrosis panel, may better reflect changes in liver fibrosis and warrant further studies. In a study of 196 patients with NAFLD, the European Liver Fibrosis panel and its simplified algorithm had good overall accuracy in diagnosing moderate and severe liver fibrosis.36 Transient elastography has also been shown to have high accuracy in detecting advanced fibrosis in patients with NAFLD.29 It remains to be seen if transient elastography may be used to monitor liver fibrosis. In addition, the cytokeratin-18 fragment is a marker of apoptotic activity and has been shown to reflect necroinflammatory activity in patients with NAFLD.37 In this study, the change in its level correlated with change in NAS.

Our study had several limitations. First, the patients were managed in a university centre. All patients received lifestyle advice and regular monitoring of metabolic factors. This might have altered the natural history of the disease. However, it would be unethical to withhold lifestyle advice in patients with high metabolic and cardiovascular risk. This study also provided evidence that improvement in metabolic factors was associated with histological benefit. Second, patients with NAFLD in a hospital clinic may have more advanced disease than those in the community. However, this cohort included a sizeable proportion of patients with simple steatosis and provided valuable data on this under-represented group. Third, the sample size of this study precluded more detailed analysis of factors associated with disease progression. Although the association between insulin resistance and adipokines and disease progression was not apparent in this cohort, this might be due to type 2 errors. Finally, liver biopsy might be limited by sampling bias. We attempted to reduce the bias by including more patients. Moreover, each histological slide was scored by two pathologists.

In conclusion, patients with NAFLD with low NAS may still develop NASH and fibrosis progression. Weight reduction is associated with non-progressive disease. All patients with NAFLD should undergo periodic assessment and lifestyle modification.

References

View Abstract

Footnotes

  • Funding The Chinese University of Hong Kong, Shatin, Hong Kong, China.

  • Competing interests None.

  • Ethics approval This study was conducted with the approval of the Chinese University Clinical Research Ethics Committee.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles

  • Digest
    Emad El-Omar Severine Vermeire Alexander Gerbes